Limnol. Oceanogr., 44(8), 1999, 1999–2004

Uptake of atmospheric CO2 by the oceans and the export of carbon into deeper waters via the biological CO2 pump is driven by the production of particulate organic matter (POM). The elemental ratios of carbon, nitrogen, and phosphorus within POM follow Redfield ratios, suggesting that the concentrations of dissolved inorganic carbon (DIC) and nutrients decrease during new production corresponding to these ratios. Subsequently, new production and the efficiency of the biological CO2 pump are usually estimated using Redfield ratios. However, our observations in the Baltic Sea and observations elsewhere show significantly greater decreases in DIC during the productive season than that predicted from the decline in nutrients with reference to Redfield ratios. Using new DIC, nutrient, and oxygen data from the Baltic Sea, we discuss this discrepancy and provide evidence that preferential recycling of nutrients fuels the productive community with nutrients. Limiting nutrients are preferentially recycled and become available for new production. These findings derived from hydrochemical bulk data confirm the picture of the microbial loop but question the common description of new production and nutrient limitation. Finally, we argue for a carbon-based efficiency estimate of the biological CO2 pump probably exceeding significantly nutrient-based estimates. Uptake of atmospheric CO2 by the oceans via the biological CO2 pump is driven by the new production of particulate organic matter (POM) and its export into deeper waters (e.g., Eppley and Peterson 1979). Phytoplankton produces POM during photosynthesis consuming dissolved inorganic carbon (DIC) and mainly nitrate (NO3) and phosphate (PO4). The elemental ratios of carbon (C), nitrogen (N), and phosphorus (P) within freshly produced POM that are found to be similar all over the world’s oceans (e.g., Copin-Monteguet and Copin-Monteguet 1983; Toggweiler 1993) are expressed by Redfield ratios, suggesting that the concentration changes of DIC and nutrients during production and remineralization of POM are in the same ratios (Redfield et al. 1963). With reference to Redfield’s idea of a coupling between DIC and nutrient concentrations and elemental ratios of POM, Dugdale and Goering (1967) deduced the widely accepted and applied description of new production and its limitation by nutrients: New production is defined as phytoplankton production based on nitrate uptake that can be converted to carbon units using the elemental ratios of POM. New production is limited by nutrient availability in the euphotic zone. Consequently, the export of CO2 into the deeper waters via the biological CO2 pump is estimated using the increases of nutrients. However, several investigations on the relationship between the elemental composition of POM and changes of nutrients over depth show only weak consistency (Shaffer 1996, and review therein). There are even strong indications that the observed DIC decrease during production of POM is decoupled from the associated NO3 decrease, because the ratio of DIC to NO3 decreases, DDIC/DNO3 is found to be significantly higher than predicted by the C/N ratio of POM. Here, we discuss some of these hydrochemical indications and refer to new DIC and NO3 data from the Baltic Sea to provide evidence that the discrepancy between observed and predicted DIC decrease is caused by preferential recycling of nutrients within the euphotic zone. This mechanism enhances the efficiency of the biological CO2 pump by allowing further netCO2 fixation, i.e., net community production, using recycled nutrients. We suggest that Redfield’s idea of fixed elemental ratios of POM is consistent with this mechanism. However, in contradiction to this mechanism are (1) the assumption that the concentration changes of DIC and nutrients would reflect the Redfield ratios, (2) the application of the definitions of new production and nutrient limitation according to Dugdale and Goering (1967) to assess net community production, and (3) the description of the biological CO2 pump with reference to nutrient concentrations and Redfield ratios.